Adaptive Learning and Personalization Within the Enterprise

Examples of Personalized Experiences

Personalization and adaptive learning also assist organizations in achieving a substantial number of different outcomes. Far from being restricted to information retrieval, they’re also helpful for application building and deployment. Trefler described an ideal experience in which an application is comprised of “a collection of processes, data, and security, but [it is] able to come to you so that you can tell it what you need and the right application will reach out to give you the right workflow to get things done.”

Reasoning models, also called thinking models, furnish vital expressions of adaptive learning based on their cyclical approach to completing tasks or solving problems. Widely deployed for workflows involving agents, these models can generate a series of outputs based on what users want until they get it right. Bradford explained: The model “comes up with ‘This is what I think you’re asking for.’ It takes that information, and if you want to think about it, it sends it back into the LLM [large language model] with both of those pieces of information. So, now the LLM knows that this is what you asked and this is what the initial pass was thinking. It’s going to take those two things, combine them together, and say, based upon this information, ‘This is the next step I’m going to do.’”

Information retrieval is still the quintessential KM application for personalized experiences. By creating 1) knowledge graphs of human users and their interests, 2) agents, which may be digital twins, for those graphs, complete with the requisite skills and tooling, and 3) respective graphs—complete with their own agents—of all the individual sources or databases, Aasman posited that “You could have agents from these various databases that understand which analyst is actually interested in this subject; let’s just send his agents this information. He doesn’t have to look for it; we’ll just give it straight to the agent, so it’ll deal with it.”

For each of these use cases, personalization systems should include a plethora of information uniquely tailored to the user’s task. Or, as Trefler put it, “The heart of your application, the heart of your business, should be a combination of the decisions you make, the workflows you execute, and the context, the case data that you collect, that you can describe, show an auditor, and do machine learning from.”

Shades of Personalization

There are several factors that organizations can use to base personalized experiences on and those involving adaptive learning. The most obvious stem from previous uses of a given application, which might include prior search results, or learning based on historic factors endemic to the user. Other facets of personalization that most organizations include pertain to “the role, the domain, and the industry you’re in,” Grout noted.

However, as various forms of adaptive learning are employed, and the personalization results from them adhere to the specific preferences of an individual, there are numerous other considerations that may be less apparent. Many of these pertain to a particular organization’s company culture—and possibly the cultural tendencies of the individual interacting with adaptive learning systems. 

As Grout mentioned, a preferred learning style may be beneficial for certain individuals. “Personalization also comes down to the medium you’re serving up information from,” he continued. “It may be the case that a shortform video sequence is the best way to serve up a piece of learning, such as a generated TikTok video style for certain demographics more than other demographics.” Age groups and geography may be germane demographics from which to create personalized experiences for users.